Phase change memory elements are programmable by the input of energy of one form or another. Most commonly, optical or electrical energy is used.
Phase change materials are materials which can be switched between generally amorphous and generally crystalline states. These materials are used in memories where electrical, optical or other energy is used to switch the material between its different states as is well known in the art.
There are a large number of patents relating to inventions made by Ovshinsky and co-workers at Energy Conversion Devices, Inc. dating from the 1960's to recent times [1-14]. There are also many other patents in this field [15-78].
Phase change memory materials can be changed between structural states of generally amorphous and generally crystalline local order, or can be set between different detectable states of local order across a continuous spectrum between completely amorphous and completely crystalline states.
Some of the materials described by the Ovshinsky patents are switchable between two detectable structural states of generally amorphous and generally crystalline local order to accommodate the storage and retrieval of single bits of encoded binary information. It is also claimed that these materials can be set at intermediate detectable levels of local order over the entire spectrum between completely amorphous and completely crystalline states. For the latter case, the intermediate detectable levels were defined as any level over the whole range local order between the completely amorphous and the completely crystalline states and was described as a “grey scale” represented by the spectrum between the completely amorphous and the completely crystalline states.
These grey scale characteristics were used to speculate that many-state phase change memories could be built exploiting a continuously variable parameter such as resistance, where separately detectable steps between maximum and minimum levels could provide multilevel logic. However, to the inventors' knowledge, no unique or physically distinguishable characteristics have yet been identified that clearly delineate between the different so-called grey scale states, other than the continuum of variation between relative “amounts” of amorphous and crystalline local order.
Furthermore, it is not clear if the continuously variable degree of local order/disorder that provides the grey scale is stable with respect to time, environmental conditions or any unexpected, undesired or parasitic fluctuations of energy that may arise, such as fluctuations in electrical, optical, pressure or thermal energy.
In an optical phase change memory where the phase change material is switched between states by the application of optical energy, the state is detectable by properties such as: index of refraction, optical absorption, optical reflectivity, or combinations thereof. Other properties that can also be detected could be changes in volume and density, through photo-expansion or photo-compaction.
In an optical phase change memory material, a laser is generally used to supply the optical energy to cause the phase change between amorphous and crystalline states. The amount of energy applied to the memory material is a function of both the power of the laser as well as the period of time that the laser pulse is applied.
Importantly though, and not widely recognised in the prior art, is the importance of the absorption coefficient of the phase change material. If the material is transparent to the laser radiation, or if the phase change material layer is too thin, then the temperature increase can be relatively small. Similarly, the thermal conductivity and heat capacity of the phase change material is important.
The crystallisation energy is also important. As defined below, the crystallisation energy is the amount of energy per unit volume needed to substantially re-crystallise an amorphous programmable volume region of the phase change memory material. If the crystallisation energy is too high, the material requires exposure to either a higher power laser pulse or a longer laser pulse in order to convert the material from the amorphous to the crystalline states. It is desirable to be able to control the crystallisation energy of a phase change memory material via the addition of one or more modifier elements. It is also desirable to increase the erasability of optical recording media. If the crystallisation energy is too low, the memory material will be unstable and information stored could be irretrievable lost.
Electrical phase change memory is capable of being electrically switched between generally amorphous and generally crystalline states for electronic memory applications. As mentioned above, it is also postulated that the material can be electrically modified between many different detectable states of local order across the continuum between completely amorphous and completely crystalline states.
That is, the electrical switching of such materials can take place between completely amorphous and completely crystalline states in a binary system, or between a larger number of incremental steps having different degrees of local order to provide a “grey scale”. Alternatively, a binary system can be contemplated which switches between two intermediate states in the continuum where one is more amorphous and less crystalline than the other.
The “grey scale” described in the Ovshinsky patents is counter-intuitive and its physical basis is unclear and has yet to be convincingly explained either by theory or experiment. The explanation given in the Ovshinsky patents is that a memory element is transferred from its high resistance state to its low resistance state through a series of sub-interval pulses and, with application of each sub-interval pulse, the resistance of the memory device does not substantially change until the total integrated duration of the sub-interval pulses is equal to or greater than a set duration. Once the final sub-interval pulse has delivered the last increment of the energy, the device is said to be transformed to the low resistance state.
In summary, while a many-state material would have clear advantages in terms of being able to store higher densities of data per unit area, it is not yet clear that a suitably stable phase change material for implementing such a many-state memory has been identified.
More generally, while phase change memories can be, and have been, made successfully, their penetration into mainstream markets is limited by several factors.
The main materials system used to date is based on GeSbTe compounds, sometimes referred to as GST, typically including one or more further elements referred to as “modifiers” or “dopants” in order to improve or otherwise adjust one or more relevant properties of the device, such as switching speed or energy.
Although GeSbTe and related phase-change compounds have generally favourable properties for phase change data storage, they have the following limitations. They provide only relatively slow electrical switching speeds. They require relatively high energy consumption, especially for writing. They could have better stability, both for long term data storage and to be stable against changes in environmental conditions, this being the case especially for multi-level devices that exploit the “grey scale”. These limitations all follow from the inherent materials properties of the phase change compound and are thus relevant for any kind of phase change memory device made from such material, whether the devices are actuated optically, electrically or otherwise.